Industrial camera supplementary light source adaptive control method and device

Abstract

The present invention relates to an adaptive control method and device for fill light sources for industrial cameras. The method comprises acquiring scene perception data from an industrial camera, performing real-time classification of the shooting scene and environmental parameter analysis based on the scene perception data to obtain scene feature information; performing image quality assessment on the scene feature information and comparing it with a preset image quality standard to obtain an image quality matching degree; performing control parameter analysis based on the image quality matching degree and the scene feature information to obtain initial fill light control parameters; acquiring image data from the industrial camera, performing brightness distribution and detail feature analysis on the image data to obtain an image analysis result; performing dynamic light source simulation on the scene feature information based on the image analysis result to obtain corresponding dynamic fill light source parameters; and comprehensively processing the initial fill light control parameters, image analysis results, and dynamic fill light source parameters to obtain a fill light source adjustment strategy. The present invention can effectively solve the problem that traditional fixed fill light methods are difficult to adapt to complex environments.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial cameras, and in particular to a method and device for adaptively controlling a supplementary light source for an industrial camera. Background Art

[0002] Industrial cameras play an increasingly important role in modern manufacturing and quality control. With the rapid development of Industry 4.0 and smart manufacturing, the requirements for industrial camera image quality are constantly increasing. However, in practical applications, the shooting environments of industrial cameras are often complex and changing, with diverse lighting conditions, which poses a significant challenge to image clarity and accuracy. Traditional fill-lighting methods for industrial cameras typically use fixed fill-lighting parameters, which are difficult to adapt to the needs of different scenes and environmental conditions. This single fill-lighting strategy can lead to unstable image quality, affecting subsequent image processing and analysis results. Therefore, how to achieve intelligent and adaptive control of fill-lighting for industrial cameras to adapt to different shooting scenes and environmental conditions and improve the stability and reliability of image quality has become an important research direction. Summary of the Invention

[0003] The main purpose of the present invention is to provide an industrial camera fill light adaptive control method and device, which can effectively solve the problem that traditional fixed fill light methods are difficult to adapt to complex environments.

[0004] To achieve the above objectives, the present invention provides an adaptive control method for fill light sources of industrial cameras, comprising:

[0005] Acquire scene perception data from an industrial camera, perform real-time classification of shooting scenes and environmental parameter analysis based on the scene perception data, and obtain scene feature information;

[0006] Performing image quality evaluation on the scene feature information and comparing it with a preset image quality standard to obtain an image quality matching degree;

[0007] Performing control parameter analysis based on the image quality matching degree and the scene feature information to obtain initial fill light control parameters;

[0008] Acquire image data from the industrial camera, perform brightness distribution and detail feature analysis on the image data, and obtain corresponding image analysis results;

[0009] Performing dynamic light source simulation on the scene feature information based on the image analysis result to obtain corresponding dynamic fill light source parameters;

[0010] The initial fill light control parameters, the image analysis results and the dynamic fill light source parameters are comprehensively processed to obtain a fill light source adjustment strategy.

[0011] Furthermore, the scene perception data of the industrial camera is obtained, and based on the scene perception data, real-time classification of the shooting scene and environmental parameter analysis are performed to obtain scene feature information, including:

[0012] Acquiring spectral data and depth data collected by the industrial camera and integrating them to obtain the scene perception data;

[0013] Performing high-order singular value decomposition on the scene perception data and extracting cross-modal correlation features to obtain a modal coupling matrix;

[0014] Performing scene construction based on the modal coupling matrix to obtain a scene semantic graph;

[0015] Performing Betti number sequence calculation on the scene semantic graph according to a preset persistent homology algorithm to obtain a scene topology feature vector;

[0016] Performing multi-scale fractal dimension analysis based on the scene topology feature vector to obtain a scene complexity spectrum;

[0017] Performing dynamic stability feature analysis based on the scene complexity spectrum to obtain scene dynamic features;

[0018] Performing symbolic processing on the dynamic features of the scene to generate a symbol sequence of the scene;

[0019] Calculating conditional entropy and mutual information based on the symbol sequence to obtain a scene uncertainty index;

[0020] The scene topology feature vector, the scene dynamic feature and the scene uncertainty index are subjected to nonlinear mapping fusion to obtain the scene feature information.

[0021] Furthermore, performing image quality assessment on the scene feature information and comparing it with a preset image quality standard to obtain an image quality matching degree includes:

[0022] Calculating the brightness entropy value of the scene feature information to obtain a scene brightness entropy value;

[0023] Performing a brightness uniformity evaluation on the scene feature information according to the scene brightness entropy value to obtain a brightness uniformity score;

[0024] Performing texture spectrum analysis on the scene feature information to obtain texture spectrum features;

[0025] Performing detail retention evaluation on the scene feature information according to the texture spectrum features to obtain a detail retention score;

[0026] Performing color saturation calculation on the scene feature information to obtain a color saturation value;

[0027] Performing a color restoration evaluation on the scene feature information according to the color saturation value to obtain a color restoration score;

[0028] Comprehensively calculating the brightness uniformity score, the detail retention score, and the color reproduction score to obtain a comprehensive image quality score;

[0029] A matching degree analysis is performed between the comprehensive image quality score and the image quality standard to obtain an image quality matching degree.

[0030] Furthermore, the performing of control parameter analysis based on the image quality matching degree and the scene feature information to obtain initial fill light control parameters includes:

[0031] Performing a nonlinear transformation of a hyperbolic tangent function on the image quality matching degree to obtain a quality matching curve;

[0032] Extracting high-order statistical features and frequency domain features from the scene feature information, and constructing a multidimensional feature space to obtain a scene feature vector;

[0033] Performing parameter analysis on the quality matching curve and the scene feature vector to obtain an initial fill light parameter set;

[0034] Iteratively calculating the initial fill light parameter set according to a preset image clarity evaluation function to obtain optimized gradient information;

[0035] The initial fill light parameter set is dynamically adjusted according to the optimization gradient information to obtain the initial fill light control parameters.

[0036] Furthermore, the acquiring of image data from the industrial camera, performing brightness distribution and detail feature analysis on the image data, and obtaining corresponding image analysis results, includes:

[0037] Performing multi-scale pyramid decomposition on the image data to obtain multi-scale frequency coefficients;

[0038] Performing energy distribution analysis on each sub-band based on the multi-scale frequency coefficients to obtain a sub-band frequency domain feature map;

[0039] Performing region value segmentation on the sub-band frequency domain feature map to obtain a significant region matrix;

[0040] Performing region division on the image data according to the salient region matrix to obtain a target region;

[0041] Performing local directional feature analysis on the target area to obtain local directional features;

[0042] Calculating the direction consistency of regional pixels based on the local direction features to obtain a detail descriptor;

[0043] Performing cluster analysis on the detail descriptors to obtain an image texture complexity index;

[0044] Performing matrix construction based on the sub-band frequency domain feature map and the texture complexity index to obtain a constructed joint distribution matrix;

[0045] Performing vector eigendecomposition on the joint distribution matrix to obtain an image matrix vector;

[0046] An image quality score is performed based on the image matrix vector and the joint distribution matrix to obtain the image analysis result.

[0047] Furthermore, the performing of dynamic light source simulation on the scene feature information based on the image analysis result to obtain corresponding dynamic fill light source parameters includes:

[0048] Performing illumination non-uniformity evaluation on the image analysis results to obtain illumination uniformity information;

[0049] Performing a virtual adjustment of the light source position on the scene feature information according to the illumination uniformity information to obtain multiple sets of virtual light source position parameters;

[0050] Performing ray tracing calculation on each set of virtual light source position parameters to obtain corresponding illumination distribution data;

[0051] Performing an object surface reflection characteristic analysis on the scene feature information according to the illumination distribution data to obtain a surface reflectivity distribution map;

[0052] Correcting the illumination distribution data based on the surface reflectance distribution map to obtain corrected illumination distribution data;

[0053] Performing light intensity optimization calculation on the corrected light distribution data to obtain optimized light intensity parameters;

[0054] generating a plurality of groups of candidate fill light solutions according to the optimized light intensity parameters and the virtual light source position parameters;

[0055] Performing image quality prediction and evaluation on each group of candidate fill light solutions to obtain corresponding image quality prediction indicators;

[0056] sorting and screening the candidate fill light solutions according to the image quality prediction index to obtain a screened fill light solution;

[0057] Dynamically converting the light source position and light intensity parameters of the screened fill light solution to obtain the dynamic fill light source parameters.

[0058] Furthermore, the comprehensive processing of the initial fill light control parameters, the image analysis results, and the dynamic fill light source parameters to obtain a fill light source adjustment strategy includes:

[0059] Performing fractal dimension analysis on the initial fill light control parameters to obtain a fill light complexity index;

[0060] Constructing an optical flow field based on the image analysis results to obtain a scene dynamic optical flow field map;

[0061] Performing tensor decomposition on the dynamic supplementary light source parameters to obtain multimodal light source features;

[0062] A topological structure is constructed based on the fill light complexity index and the scene dynamic optical flow field map to obtain an adaptive fill light network structure;

[0063] Mapping the multimodal light source characteristics onto the adaptive fill light network structure to obtain a preliminary fill light strategy;

[0064] The fill light source control pulse sequence is optimized based on the preliminary fill light strategy to obtain the fill light source adjustment strategy.

[0065] Furthermore, mapping the multimodal light source characteristics onto the adaptive fill light network structure to obtain a preliminary fill light strategy includes:

[0066] Performing feature vector decomposition on the multimodal light source features to obtain principal components of light source features;

[0067] Assigning node weights to the adaptive fill light network structure according to the principal components of the light source characteristics to obtain a weighted fill light network;

[0068] Performing graph analysis on the weighted light-filling network and extracting network topology features to obtain a network structure feature vector;

[0069] Performing a tensor product operation on the network structure feature vector and the light source feature principal component to obtain a fused feature tensor;

[0070] Performing a nonlinear transformation on the fused feature tensor to obtain a candidate set of fill light strategies;

[0071] Filtering the fill light strategy candidate set according to a preset fill light sorting standard to obtain a corresponding filtered fill light strategy set;

[0072] The screened fill light strategy set is integrated to obtain the preliminary fill light strategy.

[0073] The present invention also provides an industrial camera fill light source adaptive control device, which is applied to any of the above industrial camera fill light source adaptive control methods, comprising:

[0074] An acquisition module is used to obtain scene perception data from an industrial camera, and to perform real-time classification of shooting scenes and environmental parameter analysis based on the scene perception data to obtain scene feature information;

[0075] An analysis module is used to perform image quality evaluation on the scene feature information and compare it with a preset image quality standard to obtain an image quality matching degree;

[0076] an association module, configured to perform control parameter analysis based on the image quality matching degree and the scene feature information to obtain initial fill light control parameters;

[0077] a processing module, the processing module being used to acquire image data from the industrial camera, perform brightness distribution and detail feature analysis on the image data, and obtain corresponding image analysis results;

[0078] A control module, configured to perform dynamic light source simulation on the scene feature information based on the image analysis result to obtain corresponding dynamic fill light source parameters;

[0079] An execution module is used to comprehensively process the initial fill light control parameters, the image analysis results and the dynamic fill light source parameters to obtain a fill light source adjustment strategy.

[0080] The present invention provides an adaptive control method and device for supplementary light sources for industrial cameras, which has the following beneficial effects:

[0081] By acquiring scene perception data and performing real-time classification and environmental parameter analysis, the system can more accurately assess fill light requirements for different shooting scenarios, thereby improving fill light control precision and providing a more reliable foundation for image acquisition. By evaluating image quality based on scene feature information and comparing it against preset standards, the system enables refined management of fill light requirements for different scenarios, helping to implement on-demand fill light and avoid wasted light source resources. Control parameter analysis based on image quality matching and scene feature information ensures the fill light system operates efficiently under diverse environmental conditions, reducing unnecessary energy consumption and improving overall system efficiency. By analyzing the brightness distribution and detail characteristics of image data and performing dynamic light source simulation, a more optimized fill light strategy is developed, effectively utilizing light source resources and improving image quality stability. By comprehensively processing initial fill light control parameters, image analysis results, and dynamic fill light source parameters, the system can flexibly adjust the fill light strategy based on the characteristics and changing needs of different scenes, making the system more adaptable to diverse industrial applications and effectively addressing the difficulty of traditional fixed fill light methods in adapting to complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] Figure 1 This is a flow chart of an adaptive control method for fill light source of an industrial camera provided by the present invention;

[0083] Figure 2 This is a structural diagram of an adaptive control device for fill light source of an industrial camera provided by the present invention.

[0084] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0085] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0086] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0087] Reference Figure 1 As shown, the present invention provides an industrial camera supplementary light source adaptive control method, comprising:

[0088] Step S1: Acquire scene perception data from an industrial camera, perform real-time classification of the shooting scene and environmental parameter analysis based on the scene perception data, and obtain scene feature information;

[0089] Step S2: performing image quality assessment on the scene feature information and comparing it with a preset image quality standard to obtain an image quality matching degree;

[0090] Step S3: performing control parameter analysis based on the image quality matching degree and scene feature information to obtain initial fill light control parameters;

[0091] Step S4: Acquire image data from an industrial camera, perform brightness distribution and detail feature analysis on the image data, and obtain corresponding image analysis results;

[0092] Step S5: performing dynamic light source simulation on the scene feature information based on the image analysis results to obtain corresponding dynamic fill light source parameters;

[0093] Step S6: Comprehensively process the initial fill light control parameters, the image analysis results, and the dynamic fill light source parameters to obtain a fill light source adjustment strategy.

[0094] Based on the above steps, the detailed process is as follows:

[0095] Step S1: The industrial camera uses auxiliary equipment such as light sensors and temperature and humidity sensors to achieve comprehensive perception of the shooting environment. The light sensor accurately measures key parameters such as ambient light intensity and color temperature, while the temperature and humidity sensor monitors the physical state of the environment. The camera also records spatiotemporal information such as the shooting time and geographic location, providing basic data for subsequent scene analysis.

[0096] Real-time scene classification rapidly categorizes the current scene into different types, such as indoor / outdoor, bright / dark, and static / dynamic. This classification relies not only on the visual features of the image itself but also considers multiple dimensions, such as lighting and background. By establishing a scene feature library, the classification algorithm can be continuously optimized, improving recognition accuracy and real-time performance.

[0097] Environmental parameter analysis is an in-depth exploration and interpretation of scene perception data. It comprehensively assesses the stability, changing trends, color temperature, and color rendering of ambient lighting, and calculates the spatial distribution characteristics of the lighting. Physical parameters such as temperature and humidity are also included in the analysis, as they significantly affect image quality. This multi-dimensional, multi-level comprehensive analysis yields a more comprehensive and accurate description of the environment.

[0098] The final output of scene feature information is a standardized feature vector, which integrates scene classification results, environmental parameter analysis, and other information. This vector not only records the objective characteristics of the current scene, but also contains a deep understanding of the imaging environment.

[0099] Step S2: The core evaluation dimensions of image quality are brightness uniformity, contrast, clarity, and color reproduction. Brightness uniformity reflects the consistency of brightness across the image, contrast reflects the image's layering and detail, clarity measures image sharpness, and color reproduction reflects the authenticity and saturation of colors.

[0100] The development of pre-defined image quality standards must fully consider the needs of different application scenarios. For example, image quality requirements vary across industrial inspection, medical imaging, and security monitoring. Therefore, quality standards should be multi-tiered and adjustable. By setting weights and thresholds for each indicator, a flexible and precise evaluation system can be constructed.

[0101] Image quality assessment is implemented through intelligent prediction based on scene feature information. The likely image quality level is estimated based on scene type, environmental parameters, and other factors. A detailed quality assessment is generated by calculating scores for various indicators and performing a weighted synthesis.

[0102] The quality match calculation precisely compares the assessment results with the preset standard. The degree of match for each metric is calculated separately, ultimately yielding an overall match. Match is not just a numerical value; it quantifies the degree to which image quality deviates from the ideal. A low match indicates that imaging needs to be improved, such as with fill lighting.

[0103] Step S3: Fill light requirements analysis is a prerequisite for generating control parameters. The priority and urgency of fill light will be determined based on the quality match, and the specific requirements of the scene characteristics for fill light will be comprehensively evaluated. The difficulty and feasibility of fill light implementation vary significantly in different scenarios.

[0104] Initial parameter calculations are precisely derived based on preliminary analysis. Basic fill light requirements are calculated based on scene characteristics, and fill light intensity is then adjusted based on quality matching. Optimizing fill light timing is also crucial, taking into account the response characteristics of the light source and scene variations. The resulting parameter set should both meet theoretical expectations and be practically feasible.

[0105] Parameter feasibility verification is a critical step in ensuring the reliability of the fill light control solution. We thoroughly check whether the generated parameters are within the device's permitted range, evaluate parameter stability, and predict the fill light effect. Only parameters that pass rigorous verification are actually implemented, which is a key guarantee for the reliability of the fill light system.

[0106] Step S4: The industrial camera captures high-precision, high-resolution raw image data, ensuring its integrity and authenticity. The data acquisition process must consider key technical indicators such as the image sensor's dynamic range and signal-to-noise ratio to ensure input data quality. A standardized data preprocessing process is established to perform necessary denoising and correction on the raw images.

[0107] Brightness distribution analysis is a core component of image quality assessment. Image processing algorithms are employed to perform a detailed, multi-dimensional, and multi-scale analysis of image brightness. Histogram analysis and regional brightness mapping techniques are used to accurately characterize the spatial distribution of image brightness. This focus is on brightness uniformity, gradient variations, and brightness contrast in key areas, providing crucial information for subsequent fill-lighting decisions.

[0108] Detail feature analysis meticulously mines deep-level image information. It extracts key features such as texture, edges, and grain. The analysis spans multiple scales and levels of abstraction, focusing not only on macroscopic structures but also delving deeply into microscopic details. Through feature mapping and feature importance assessment, we fully understand the structural complexity and information richness of the image.

[0109] The output of image analysis is a multi-dimensional feature vector. It comprehensively reflects key information about the image, including brightness distribution, detail features, and texture structure. This vector not only provides a quantitative description of the image but also bridges image perception and fill-light control. It provides precise input for subsequent dynamic light source simulation and fill-light strategy development.

[0110] Step S5: Through physical optics simulation and computer rendering technology, we can accurately predict the lighting effects under different fill-light schemes. The simulation process considers multiple factors such as light source type, lighting angle, and reflection characteristics, striving to achieve a lighting reconstruction that is closest to the real environment.

[0111] Dynamic association of scene feature information is key to the simulation process. Image analysis results are deeply integrated with previously acquired scene feature information to establish a dynamic, real-time scene optical model. Machine learning algorithms continuously optimize the accuracy and adaptability of light source simulation. This model will be able to rapidly respond to scene changes, enabling intelligent prediction and simulation of lighting environments.

[0112] Generating dynamic fill light parameters is a multi-objective optimization problem. An intelligent algorithm generates optimal fill light parameters, taking into account multiple constraints, including image quality, energy consumption, and light source lifespan. These parameters encompass multiple dimensions, including fill light intensity, color temperature, and spatial distribution. The optimization goal is to achieve efficient and energy-efficient fill light operation while maintaining image quality.

[0113] Verifying parameter accuracy is a critical step in ensuring the reliability of simulation results. Through various methods, including model comparison and experimental verification, we will comprehensively evaluate the scientific and feasibility of the dynamic fill light parameters. This process is not only a technical test of the parameters, but also a continuous optimization and improvement of the entire light source simulation method.

[0114] Step S6: Comprehensively process the initial fill light control parameters, the image analysis results, and the dynamic fill light source parameters to obtain a fill light source adjustment strategy.

[0115] Comprehensive processing is the integration and optimization phase of the entire fill light control solution. It integrates the initial control parameters, image analysis results, and dynamic fill light parameters obtained earlier in a multi-dimensional and multi-level manner. The comprehensive processing results are obtained by prioritizing and collaboratively optimizing various parameters.

[0116] Developing a supplemental light source adjustment strategy is a dynamic, adaptive process. Based on comprehensive processing results, a multi-layered, evolvable adjustment framework is constructed. This strategy encompasses not only static parameter settings but also dynamic adjustment mechanisms based on real-time feedback. Through closed-loop control and self-learning algorithms, intelligent adaptation to complex and changing environments is achieved.

[0117] The present invention provides an adaptive control method for fill light sources for industrial cameras. By acquiring scene perception data and performing real-time classification and environmental parameter analysis, it can more accurately assess fill light requirements for different shooting scenes, thereby improving fill light control precision and providing a more reliable foundation for image acquisition. By performing image quality assessment on scene feature information and comparing it with preset standards, it enables refined management of fill light requirements for different scenes, facilitating on-demand fill light and avoiding waste of light source resources. Control parameter analysis based on image quality matching and scene feature information ensures that the fill light system operates efficiently under different environmental conditions, reducing unnecessary energy consumption and improving the overall operating efficiency of the system. By analyzing the brightness distribution and detail characteristics of image data and performing dynamic light source simulation, a more reasonable fill light strategy is developed, effectively utilizing light source resources and improving image quality stability. By comprehensively processing initial fill light control parameters, image analysis results, and dynamic fill light source parameters, the fill light strategy can be flexibly adjusted according to the characteristics and changing needs of different scenes, making the system more adaptable to diverse industrial application scenarios and effectively addressing the difficulty of traditional fixed fill light methods in adapting to complex environments.

[0118] In one embodiment, scene perception data from an industrial camera is acquired, and real-time classification of the shooting scene and environmental parameter analysis are performed based on the scene perception data to obtain scene feature information, including:

[0119] Industrial cameras simultaneously collect spectral data and three-dimensional depth information of a scene. During data acquisition, the spectral sensor captures electromagnetic spectrum characteristics in the wavelength range of 380-780nm, while the depth sensor uses structured light or time-of-flight (ToF) technology to accurately measure the spatial position and distance of objects in the scene.

[0120] During the data integration phase, high-order singular value decomposition (SVD) techniques are used to convert spectral and depth data into a cross-modal correlation feature matrix. This matrix effectively reduces data dimensionality and extracts key semantic information through feature decoupling and information compression. Information theory and manifold learning constraints are introduced during the construction of the modal coupling matrix to ensure robust feature representation and information integrity.

[0121] The scene semantic graph maps the modal coupling matrix to a set of topologically connected semantic nodes. Each node represents a key semantic feature in the scene, and the connection weights between nodes reflect the strength of the association between features. During the semantic graph generation process, graph structure regularization constraints are introduced to enhance the structural consistency of the graph representation.

[0122] The topological feature vectors are calculated using the persistent homology algorithm to form a Betti number sequence, which reflects the topological complexity of the scene. Based on algebraic topology theory, the persistent homology method can capture the multi-scale geometric and topological characteristics of the data and quantitatively describe the scene structure. During the calculation process, a scale parameter threshold is set to control the granularity of feature extraction.

[0123] Multiscale fractal dimension analysis employs the theory of generalized fractal dimension to quantify scene complexity at different scales and angles. Self-similarity and scale invariance constraints are incorporated into the analysis to ensure the stability and representativeness of the complexity spectrum. The scene complexity spectrum not only reflects spatial structural characteristics but also reveals the dynamic patterns of the scene.

[0124] Dynamic feature analysis is based on nonlinear dynamics theory and calculates the stability indicators of the scenario through phase space reconstruction and state transition matrices. The analysis considers the nonlinear coupling and chaotic characteristics of the system, and sets criteria such as the Lyapunov exponent and entropy complexity to characterize the dynamic evolution characteristics of the scenario.

[0125] Symbolization converts continuous dynamic features into discrete symbol sequences, using fuzzy clustering and information quantization methods. Information entropy and fuzzy membership constraints are introduced during the symbol sequence generation process to ensure the integrity and discriminability of the information represented by the symbols.

[0126] Conditional entropy and mutual information calculations reveal the uncertainty and information complexity of a scenario. Based on the principles of probability distribution and information theory, the calculation process sets an information entropy threshold and mutual information criterion to quantify the randomness and structure of the scenario.

[0127] The topological feature vectors, dynamic features and uncertainty indicators are fused through nonlinear mapping to generate comprehensive scene feature information.

[0128] This embodiment achieves precise and intelligent control of the fill light source for industrial cameras through cross-modal data fusion and complex scene analysis. It can significantly improve the adaptability of the light source and effectively meet the lighting requirements of different complex scenes. Through continuous coherence and multi-scale fractal dimension analysis, the topological structure and dynamic characteristics of the scene can be accurately captured, overcoming the limitations of traditional methods in environmental perception. By introducing symbolic processing and information theory analysis, the robustness of scene feature extraction and the efficiency of information utilization are significantly improved. Compared with traditional fill light solutions, this embodiment can more intelligently adjust the intensity and spectral characteristics of the light source, significantly improve the quality and consistency of industrial image acquisition, and provide more reliable technical support for precision industrial detection.

[0129] In one embodiment, performing image quality assessment on scene feature information and comparing it with a preset image quality standard to obtain an image quality matching degree includes:

[0130] Perform image quality assessment on the input scene feature image. The image's brightness entropy is calculated using an entropy formula. This entropy reflects the uniformity of the image's grayscale distribution. Based on the entropy value, a 0-100 score system is used to quantitatively assess the image's brightness uniformity. The closer the entropy value is to an ideal uniform distribution, the higher the brightness uniformity score.

[0131] Texture spectrum analysis is performed on the image, using a Fast Fourier Transform (FFT) to calculate the spectral energy distribution, focusing on high-, mid-, and low-frequency components. A detail retention score is calculated by evaluating the retention rate of high-frequency components. A higher retention rate indicates more complete preservation of image detail information, resulting in a higher score.

[0132] Convert the image to HSV color space and calculate the saturation histogram and average saturation value. Based on the saturation value, the image's color reproduction is evaluated using a 0-100 score system. The closer the saturation value is to the ideal range, the higher the color reproduction score.

[0133] In the comprehensive image quality scoring phase, brightness uniformity (weight 0.3), detail retention (weight 0.4), and color reproduction (weight 0.3) are weighted according to preset weights to obtain a comprehensive image quality score.

[0134] Finally, the comprehensive image quality score is compared with a preset image quality threshold (e.g., 85 points). The matching percentage between the score and the standard is calculated, and the relevant parameters of the fill light source are adjusted based on the matching feedback, achieving adaptive fill light control for industrial scenes.

[0135] In the specific implementation, the following scoring rules are set:

[0136] Brightness uniformity score:

[0137] Entropy value > 4.5: 90-100 points, entropy value 3-4.5: 70-89 points, entropy value < 3: 0-70 points.

[0138] Detail Preservation Rating:

[0139] High-frequency component retention rate > 80%: 90-100 points, retention rate 60-80%: 70-89 points,

[0140] Retention rate <60%: 0-70 points.

[0141] Color reproduction score:

[0142] Saturation 0.6-1.0: 90-100 points, saturation 0.3-0.6: 70-89 points, saturation <0.3: 0-70 points.

[0143] In one embodiment, control parameter analysis is performed based on image quality matching and scene feature information to obtain initial fill light control parameters, including:

[0144] Intelligent adaptive control of supplemental light sources is achieved through in-depth analysis of image quality matching and scene feature information. Image quality matching is assessed based on the similarity between the original image and an ideal reference image, using the Structural Similarity Index (SSIM) as a quantitative measure of matching. This matching value is then transformed nonlinearly using a hyperbolic tangent function to enhance feature sensitivity and discrimination.

[0145] Scene feature information extraction utilizes a multidimensional feature space construction strategy, beginning with high-order statistical feature extraction. The image grayscale histogram is calculated to extract high-order moment features such as skewness and kurtosis, reflecting the complexity and concentration of the image's grayscale distribution. The variance and standard deviation of the pixel grayscale distribution are also calculated to quantify image contrast and brightness variations. Frequency domain feature extraction utilizes a Fourier transform to analyze the spectral energy distribution and extract texture and structural features. The high-order statistical features and frequency domain features are then combined to form a multidimensional scene feature vector V, laying the foundation for subsequent parameter analysis.

[0146] The initial fill light parameter set is generated based on the mapping relationship between the quality matching curve Q and the scene feature vector V. Parameters such as fill light intensity, color temperature, and strobe frequency are determined according to predefined rules. The mapping model considers factors such as scene brightness, contrast, and texture complexity, and obtains the initial fill light parameter set P through nonlinear mapping in a multidimensional feature space. For example, when the brightness variance in the scene feature vector V is large, the fill light intensity is increased; when the frequency domain features show a high frequency component, the color temperature is adjusted to improve image clarity.

[0147] Furthermore, feature points are extracted from the quality matching curve, and the frequency domain feature parameters of the curve are extracted through Fourier transform. Principal component analysis is performed on the scene feature vector to construct the feature space after dimensionality reduction.

[0148] A multidimensional parameter mapping model is established based on the frequency-domain characteristic parameters of the quality matching curve and the principal components of the scene feature vector. Parameter weights are calculated for this multidimensional parameter mapping model to determine the importance of each feature dimension. Based on the parameter weights, the initial fill light parameter set is fitted using the least squares method to obtain an initial solution. Constraints are then checked on this initial fill light parameter set, and parameters that do not meet the preset constraints are eliminated.

[0149] Among them, feature points are extracted from the quality matching curve, and frequency domain feature parameters of the curve are extracted through Fourier transform, including:

[0150] The quality matching curve is subjected to discrete Fourier transform to obtain frequency domain spectrum coefficients, the amplitude and phase information of the frequency domain spectrum coefficients are extracted, and the characteristic parameter vector is constructed based on the amplitude and phase of the spectrum coefficients.

[0151] Perform principal component analysis on the scene feature vector to construct a feature space after dimensionality reduction, including:

[0152] Calculate the covariance matrix of the scene eigenvector, perform eigenvalue decomposition on the covariance matrix, select the main eigencomponents according to the size of the eigenvalues, and reconstruct the feature space with the main eigencomponents.

[0153] Parameter weight calculation based on multi-dimensional parameter mapping model, including:

[0154] Construct a parameter sensitivity matrix and calculate the information gain of each parameter through information entropy.

[0155] Iterative image clarity optimization uses a preset evaluation function E(P). Common metrics include image entropy and gradient variance. Image entropy reflects the amount of information in an image, while gradient variance reflects the clarity of edge details. Gradients are calculated for the initial parameter set P, analyzing the sensitivity of the evaluation function E(P) to each parameter. Partial derivatives are calculated to obtain the optimized gradient information G, which guides parameter adjustments. Gradient descent is a commonly used optimization strategy. The update formula is P' = P - α * G, where α is the learning rate, which controls the step size of each iteration.

[0156] The initial fill light parameter set is dynamically adjusted based on the optimized gradient information. During the iteration process, image quality changes are monitored in real time, and the fill light control parameters are continuously optimized. When the gradient converges or the preset number of iterations is reached, the optimal fill light control parameter set is output. In practical applications, the nonlinear transformation coefficients, feature extraction algorithms, and optimization strategies need to be adjusted according to the specific industrial scenarios to adapt to different image acquisition environments.

[0157] This embodiment uses the Structural Similarity Index (SSIM) to assess image quality and combines multidimensional scene feature analysis with high-order statistical features and frequency domain features to achieve precise adjustment of the fill light source. Nonlinear transformations enhance feature sensitivity, ensuring that dynamic adjustment of fill light parameters more closely matches actual scene requirements, improving image clarity and detail. Principal component analysis and the construction of a parameter sensitivity matrix make the weight calculation of feature dimensions more scientific and rational, effectively eliminating parameters that do not meet preset conditions and optimizing the fill light control strategy. An iterative image clarity optimization mechanism is introduced, using an evaluation function to monitor image quality changes in real time, ensuring continuous adjustment and optimization of fill light parameters and ultimately outputting the optimal parameter set. This approach not only improves the adaptability and flexibility of image acquisition but also reduces image quality fluctuations caused by environmental changes, providing a more stable and efficient solution for industrial applications.

[0158] In one embodiment, image data from an industrial camera is acquired, and brightness distribution and detail feature analysis are performed on the image data to obtain corresponding image analysis results, including:

[0159] Image data undergoes multi-scale pyramid decomposition, a process that utilizes a wavelet transform algorithm to decompose the original image into frequency coefficients at different scales and orientations. This decomposition involves low-frequency and high-frequency subbands, where the low-frequency subbands preserve the overall image structure, while the high-frequency subbands capture image details and texture features. For each subband, its energy distribution is calculated to generate a subband frequency domain feature map. This step helps quantify the information characteristics of the image at different scales.

[0160] During the region analysis phase, the subband frequency domain feature maps are segmented by region values. Salient regions within the image are identified by setting thresholds and using a region growing algorithm. The salient region matrix identifies representative image regions that may contain key objects or important texture information. Based on the salient region matrix, the image is divided into distinct target regions, each representing distinct image semantic content.

[0161] Local directional feature analysis is a key step in image understanding. Local directional gradients are calculated for the target region, and the local directional feature vector is constructed by analyzing the gradient direction and intensity of each pixel. This step captures the directional characteristics of the image texture and provides a foundation for subsequent texture complexity assessment.

[0162] To quantify the texture complexity of an image, pixel-wise consistency calculations are performed on local directional features to generate a detail descriptor. Using clustering methods (such as K-means clustering), the detail descriptor maps directional features to discrete complexity levels, resulting in a texture complexity index for the image. This index reflects the uniformity and structural complexity of the image texture.

[0163] During the matrix construction phase, the joint distribution matrix is ​​constructed by combining the subband frequency domain feature maps and the texture complexity index. This matrix construction adheres to the principle of maximizing information entropy and captures the multidimensional information characteristics of the image through feature fusion. Vector eigendecomposition is performed on the joint distribution matrix to extract the image's primary information eigenvectors.

[0164] Image quality is scored based on the image matrix vector and the joint distribution matrix. It considers multiple dimensions such as brightness distribution, texture complexity, and frequency domain features, and outputs a comprehensive image quality evaluation index.

[0165] This embodiment significantly improves the image quality of industrial cameras under complex lighting conditions through multi-scale image analysis technology. The use of pyramid decomposition and sub-band frequency domain feature analysis can accurately capture the details and texture information in the image, thereby effectively identifying salient areas. This process ensures that under different ambient lighting conditions, the industrial camera can automatically adjust the supplementary light source, optimize the image acquisition effect, and improve the clarity and recognizability of the image. The analysis of local directional features and the calculation of detail descriptors make the assessment of image texture complexity more accurate, thereby improving the reliability of image quality scoring. By constructing a joint distribution matrix and comprehensively considering frequency domain features and texture complexity, a comprehensive understanding of image information is achieved, providing a scientific basis for the adaptive adjustment of supplementary light sources.

[0166] In one embodiment, dynamic light source simulation is performed on scene feature information based on image analysis results to obtain corresponding dynamic fill light source parameters, including:

[0167] A comprehensive lighting uniformity assessment is performed on the input image, using algorithms such as pixel brightness standard deviation and histogram comparison to calculate the lighting differences between different areas of the image. If the lighting uniformity is detected to be below a preset threshold (for example, 0.6), a virtual adjustment process for the light source position is triggered.

[0168] Virtual light source position parameters are generated based on scene geometry and illumination distribution relationships. A spatial mapping algorithm is used to construct multiple sets of candidate light source positions. Each set of position parameters is combined with scene depth information and occlusion relationships, and the illumination distribution data is calculated using a Monte Carlo ray tracing algorithm. The ray tracing process considers the reflective properties of surface materials, simulating the propagation and attenuation of light on different surfaces to generate highly accurate illumination distribution maps.

[0169] Surface reflectivity analysis combines image textures with material recognition algorithms to construct a reflectivity distribution map for each surface object in the scene. This analysis involves calculating texture gradients, classifying materials, and extracting reflectivity parameters to identify the optical properties of different materials. Based on this reflectivity distribution map, the original illumination distribution data is accurately corrected to compensate for the complex influence of materials on light propagation.

[0170] Light intensity optimization calculations incorporate information theory, using entropy analysis and visual saliency algorithms to assess the fidelity and visual comfort of image information at varying light intensities. The optimization goal is to maximize image contrast and clarity while preserving image detail. The calculation constructs a multidimensional optimization objective function, comprehensively considering metrics such as brightness uniformity, color saturation, and texture detail preservation.

[0171] Candidate fill-light solutions are comprehensively evaluated using an image quality prediction model. This deep learning-based image quality assessment network integrates objective metrics such as peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) with subjective perceptual scores. The evaluation algorithm generates multi-dimensional quality prediction indicators for each set of fill-light solutions and employs a weighted ranking strategy to select the optimal fill-light solution.

[0172] The dynamic fill light parameter generation process involves a smooth mapping of light source position and intensity. An interpolation algorithm is used to achieve a continuous parameter transition, avoiding sudden changes in illumination. This parameter conversion takes into account scene dynamics and the perceptual characteristics of the human eye, ensuring a natural and smooth fill light process. This entire method implements intelligent, adaptive light source simulation based on image analysis, significantly improving image visual quality and information readability.

[0173] This embodiment solves the problem of local overexposure or dark areas caused by the fixed position of the light source in the traditional fill light solution by evaluating the illumination uniformity of the image and adjusting the position of the virtual light source. The reflectivity distribution map is constructed using a micro-surface reflection model and a material recognition algorithm, which overcomes the limitation of conventional fill light technology that ignores material characteristics and makes the fill light effect closer to the real scene. The light intensity parameters are optimized through information theory and visual perception models, which improves the overall visual comfort of the picture while ensuring image details. The image quality prediction model based on deep learning is used to evaluate and screen the fill light scheme, avoiding the blindness of the fill light parameter selection in the traditional method. Finally, the dynamic adjustment of the illumination parameters is achieved through a smooth mapping algorithm, which eliminates the sudden change of light intensity during the fill light process and improves the naturalness and consistency of the fill light effect. The overall solution significantly improves the intelligence level and adaptability of scene fill light and enhances image quality.

[0174] In one embodiment, the initial fill light control parameters, image analysis results, and dynamic fill light source parameters are comprehensively processed to obtain a fill light source adjustment strategy, including:

[0175] When performing fractal dimension analysis on the initial fill light control parameters, the fill light complexity index is obtained by calculating the box count dimension and information dimension of the fill light parameters. The box count dimension reflects the spatial distribution characteristics of the fill light parameters, while the information dimension represents the information entropy distribution characteristics of the parameters. Specifically, the fill light control parameter space is divided into several equal-sized grids. The number of parameter points in each grid is counted, and the dimension value is calculated using fractal theory to ultimately obtain an index value reflecting the complexity of the fill light control.

[0176] In the process of constructing the optical flow field, pixel displacement vectors between adjacent frames are calculated based on image analysis results. The Horn-Schunck algorithm is used to establish the optical flow constraint equation, which is then solved using a variational method to obtain the motion vector field for the entire scene. The optical flow field calculation must satisfy the assumptions of constant image grayscale and continuous motion, ultimately generating an optical flow map that represents the dynamic characteristics of the scene.

[0177] When performing tensor decomposition on dynamic supplemental light parameters, parameters such as brightness, color temperature, and illumination angle are organized into a third-order tensor, which is then decomposed into the product of a kernel tensor and a factor matrix. Setting an appropriate rank parameter during the decomposition process ensures that the extracted features have good expressive power, ultimately resulting in a feature representation that reflects the multidimensional properties of the supplemental light source.

[0178] During topology construction, the fill-light complexity index determines the number of network nodes, and the scene's dynamic optical flow field graph determines the connections between nodes. This allows for the construction of an adaptive fill-light control network. The network structure must meet connectivity and stability requirements, and the connection weights between nodes are dynamically adjusted based on the gradient information of the optical flow field.

[0179] The feature mapping process uses a nonlinear projection method to map the multimodal light source features onto network nodes. The activation state of the node reflects the value of the corresponding fill light parameter. A nonlinear function is used as the mapping function to ensure the smoothness of the feature projection, thereby obtaining a preliminary fill light control strategy.

[0180] In the pulse sequence optimization for fill light control, pulse waveforms and timing parameters are designed based on the preliminary fill light strategy. The optimization objectives include minimizing energy consumption, ensuring lighting uniformity, and avoiding stroboscopic effects. A genetic algorithm is used to determine the optimal pulse sequence parameters, ultimately forming a complete fill light control strategy.

[0181] This embodiment achieves accurate quantification of the spatial distribution characteristics of the fill light parameters by performing fractal dimension analysis on the initial fill light control parameters, effectively improving the accuracy of the fill light control. By constructing a dynamic optical flow field of the scene, the motion characteristics in the scene are accurately captured, providing reliable dynamic information support for the formulation of the fill light strategy. The adaptive fill light network constructed based on the fill light complexity index and the dynamic characteristics of the scene has strong environmental adaptability and can flexibly adjust the fill light strategy according to scene changes. Through nonlinear feature mapping and pulse sequence optimization, energy consumption is minimized while ensuring the lighting effect, effectively avoiding the impact of the stroboscopic effect on the image quality. The entire control method fully considers multiple key factors in the fill light process, significantly improving the fill light control effect and image acquisition quality of industrial cameras in complex scenes.

[0182] In one embodiment, the multimodal light source characteristics are mapped to the adaptive fill light network structure to obtain a preliminary fill light strategy, including:

[0183] When performing eigenvector decomposition on multimodal light source characteristics, principal component analysis (PCA) is used to decompose the light source characteristic matrix into the product of an eigenvector matrix and a diagonal eigenvalue matrix. Eigenvalues ​​are arranged from largest to smallest, and eigenvectors with a cumulative contribution rate of 95% are selected as the principal components of the light source characteristics.

[0184] When assigning weights to the adaptive fill-light network structure, the correlation coefficient between nodes is calculated based on the principal components of the light source characteristics and used as the weight value of the connecting edge. The weight assignment must meet the network connectivity requirement, and there must be at least one valid path between any two nodes.

[0185] During the graph analysis phase, spectral clustering is performed on the weighted light-filling network to extract the network's topological features. The network's Laplace matrix eigenvalues ​​and eigenvectors are calculated to obtain the network's structural eigenvector. This eigenvector contains topological information such as the network's connectivity and clustering coefficient.

[0186] The fused feature tensor is obtained by performing a tensor product operation on the network structure feature vector and the principal component of the light source feature. The tensor product operation maintains the correlation between the two types of features, forming a high-dimensional feature representation. During this operation, the tensor dimensions must be matched.

[0187] When performing nonlinear transformations on the fused feature tensor, the ReLU activation function is used to implement nonlinear mapping of features and generate a candidate set of fill-in strategies. Nonlinear transformations enhance the expressive power of features and make fill-in strategies more adaptable.

[0188] The selection of fill-light strategies is based on preset ranking criteria, including indicators such as illumination uniformity, energy efficiency, and spot distribution. A multi-objective scoring system is applied to candidate strategies, and strategies with a comprehensive score in the top 30% are selected as the selected fill-light strategy set.

[0189] In the integration phase, a weighted average is performed on the selected strategies in the fill-light strategy set. The weight coefficient is determined by the score of each strategy. The integrated strategy must meet constraints such as light intensity and energy consumption to form the final preliminary fill-light strategy.

[0190] This embodiment achieves effective extraction of light source features and optimized configuration of network structure by decomposing feature vectors of multimodal light source features and mapping them to the adaptive fill-light network structure. The principal components of light source features are extracted based on the principal component analysis method, which retains key feature information and reduces data redundancy. The spectral clustering method is used to extract network topology features, effectively capturing the structural characteristics of the network. Feature fusion is achieved through tensor product operations, which enhances the feature expression ability and makes the fill-light strategy more adaptable. In the strategy screening process, a multi-objective scoring mechanism is adopted to ensure the illumination uniformity and energy efficiency of the screened fill-light strategy. Finally, the strategies are integrated by weighted averaging, and under the premise of meeting the constraints, a fill-light solution that takes into account both illumination effects and system performance is formed, thereby improving the imaging quality and system operation efficiency of industrial cameras in different scenarios.

[0191] Reference Figure 2 As shown, the present invention also provides an industrial camera supplementary light source adaptive control device, which is applied to any of the above industrial camera supplementary light source adaptive control methods, including:

[0192] The acquisition module is used to obtain scene perception data from industrial cameras, perform real-time classification of shooting scenes and environmental parameter analysis based on the scene perception data, and obtain scene feature information;

[0193] An analysis module is used to evaluate the image quality of scene feature information and compare it with a preset image quality standard to obtain an image quality matching degree;

[0194] An association module is used to analyze control parameters based on image quality matching and scene feature information to obtain initial fill light control parameters;

[0195] The processing module is used to obtain image data from industrial cameras, perform brightness distribution and detail feature analysis on the image data, and obtain corresponding image analysis results;

[0196] A control module is used to perform dynamic light source simulation on scene feature information based on image analysis results to obtain corresponding dynamic fill light source parameters;

[0197] The execution module is used to comprehensively process the initial fill light control parameters, image analysis results and dynamic fill light source parameters to obtain a fill light source adjustment strategy.

[0198] The present invention provides an adaptive fill light control device for industrial cameras. By acquiring scene perception data and performing real-time classification and environmental parameter analysis, it can more accurately assess fill light requirements for different shooting scenes, thereby improving fill light control precision and providing a more reliable foundation for image acquisition. By performing image quality assessment on scene feature information and comparing it with preset standards, it enables refined management of fill light requirements for different scenes, helping to implement on-demand fill light and avoid wasting light source resources. Control parameter analysis based on image quality matching and scene feature information ensures that the fill light system can operate efficiently under different environmental conditions, reducing unnecessary energy consumption and improving the overall operating efficiency of the system. By analyzing the brightness distribution and detail characteristics of image data and performing dynamic light source simulation, a more reasonable fill light strategy is formulated, effectively utilizing light source resources and improving image quality stability. By comprehensively processing initial fill light control parameters, image analysis results, and dynamic fill light parameters, the fill light strategy can be flexibly adjusted according to the characteristics and changing needs of different scenes, making the system more adaptable to diverse industrial application scenarios and effectively solving the problem that traditional fixed fill light methods are difficult to adapt to complex environments.

[0199] It should be noted that, those skilled in the art will clearly understand that, for the sake of convenience and brevity of description, the specific working processes of the above-described system and each module can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0200] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for adaptively controlling supplementary light sources for industrial cameras, characterized in that: include: Acquire scene perception data from an industrial camera, perform real-time classification of shooting scenes and environmental parameter analysis based on the scene perception data, and obtain scene feature information; Performing image quality evaluation on the scene feature information and comparing it with a preset image quality standard to obtain an image quality matching degree; Performing control parameter analysis based on the image quality matching degree and the scene feature information to obtain initial fill light control parameters; Acquire image data from the industrial camera, perform brightness distribution and detail feature analysis on the image data, and obtain corresponding image analysis results; Performing dynamic light source simulation on the scene feature information based on the image analysis result to obtain corresponding dynamic fill light source parameters; Comprehensively processing the initial fill light control parameters, the image analysis results, and the dynamic fill light source parameters to obtain a fill light source adjustment strategy; Based on the scene perception data, real-time classification of the shooting scene and environmental parameter analysis are performed to obtain scene feature information, including: Acquiring spectral data and depth data collected by the industrial camera and integrating them to obtain the scene perception data; Performing high-order singular value decomposition on the scene perception data and extracting cross-modal correlation features to obtain a modal coupling matrix; Performing scene construction based on the modal coupling matrix to obtain a scene semantic graph; Performing Betti number sequence calculation on the scene semantic graph according to a preset persistent homology algorithm to obtain a scene topology feature vector; Performing multi-scale fractal dimension analysis based on the scene topology feature vector to obtain a scene complexity spectrum; Performing dynamic stability feature analysis based on the scene complexity spectrum to obtain scene dynamic features; Performing symbolic processing on the dynamic features of the scene to generate a symbol sequence of the scene; Calculating conditional entropy and mutual information based on the symbol sequence to obtain a scene uncertainty index; The scene topology feature vector, the scene dynamic feature and the scene uncertainty index are subjected to nonlinear mapping fusion to obtain the scene feature information.

2. The method for adaptively controlling the supplementary light source of an industrial camera according to claim 1, characterized in that: The performing image quality assessment on the scene feature information and comparing it with a preset image quality standard to obtain an image quality matching degree includes: Calculating the brightness entropy value of the scene feature information to obtain a scene brightness entropy value; Performing a brightness uniformity evaluation on the scene feature information according to the scene brightness entropy value to obtain a brightness uniformity score; Performing texture spectrum analysis on the scene feature information to obtain texture spectrum features; Performing detail retention evaluation on the scene feature information according to the texture spectrum features to obtain a detail retention score; Performing color saturation calculation on the scene feature information to obtain a color saturation value; Performing a color restoration evaluation on the scene feature information according to the color saturation value to obtain a color restoration score; Comprehensively calculating the brightness uniformity score, the detail retention score, and the color reproduction score to obtain a comprehensive image quality score; A matching degree analysis is performed between the comprehensive image quality score and the image quality standard to obtain an image quality matching degree.

3. The method for adaptively controlling the supplementary light source of an industrial camera according to claim 1, wherein: The performing control parameter analysis based on the image quality matching degree and the scene feature information to obtain initial fill light control parameters includes: Performing a nonlinear transformation of a hyperbolic tangent function on the image quality matching degree to obtain a quality matching curve; Extracting high-order statistical features and frequency domain features from the scene feature information, and constructing a multidimensional feature space to obtain a scene feature vector; Performing parameter analysis on the quality matching curve and the scene feature vector to obtain an initial fill light parameter set; Iteratively calculating the initial fill light parameter set according to a preset image clarity evaluation function to obtain optimized gradient information; The initial fill light parameter set is dynamically adjusted according to the optimization gradient information to obtain the initial fill light control parameters.

4. The method for adaptively controlling the supplementary light source of an industrial camera according to claim 1, wherein: The acquiring of image data from the industrial camera, performing brightness distribution and detail feature analysis on the image data, and obtaining corresponding image analysis results includes: Performing multi-scale pyramid decomposition on the image data to obtain multi-scale frequency coefficients; Performing energy distribution analysis on each sub-band based on the multi-scale frequency coefficients to obtain a sub-band frequency domain feature map; Performing region value segmentation on the sub-band frequency domain feature map to obtain a significant region matrix; Performing region division on the image data according to the salient region matrix to obtain a target region; Performing local directional feature analysis on the target area to obtain local directional features; Calculating the direction consistency of regional pixels based on the local direction features to obtain a detail descriptor; Performing cluster analysis on the detail descriptors to obtain an image texture complexity index; Performing matrix construction based on the sub-band frequency domain feature map and the texture complexity index to obtain a constructed joint distribution matrix; Performing vector eigendecomposition on the joint distribution matrix to obtain an image matrix vector; An image quality score is performed based on the image matrix vector and the joint distribution matrix to obtain the image analysis result.

5. The method for adaptively controlling the supplementary light source of an industrial camera according to claim 1, wherein: The performing dynamic light source simulation on the scene feature information based on the image analysis result to obtain corresponding dynamic fill light source parameters includes: Performing illumination non-uniformity evaluation on the image analysis results to obtain illumination uniformity information; Performing a virtual adjustment of the light source position on the scene feature information according to the illumination uniformity information to obtain multiple sets of virtual light source position parameters; Performing ray tracing calculation on each set of virtual light source position parameters to obtain corresponding illumination distribution data; Performing an object surface reflection characteristic analysis on the scene feature information according to the illumination distribution data to obtain a surface reflectivity distribution map; Correcting the illumination distribution data based on the surface reflectance distribution map to obtain corrected illumination distribution data; Performing light intensity optimization calculation on the corrected light distribution data to obtain optimized light intensity parameters; generating a plurality of groups of candidate fill light solutions according to the optimized light intensity parameters and the virtual light source position parameters; Performing image quality prediction and evaluation on each group of candidate fill light solutions to obtain corresponding image quality prediction indicators; sorting and screening the candidate fill light solutions according to the image quality prediction index to obtain a screened fill light solution; Dynamically converting the light source position and light intensity parameters of the screened fill light solution to obtain the dynamic fill light source parameters.

6. The method for adaptively controlling the supplementary light source of an industrial camera according to claim 1, characterized in that: The comprehensive processing of the initial fill light control parameters, the image analysis results, and the dynamic fill light source parameters to obtain a fill light source adjustment strategy includes: Performing fractal dimension analysis on the initial fill light control parameters to obtain a fill light complexity index; Constructing an optical flow field based on the image analysis results to obtain a scene dynamic optical flow field map; Performing tensor decomposition on the dynamic supplementary light source parameters to obtain multimodal light source features; A topological structure is constructed based on the fill light complexity index and the scene dynamic optical flow field map to obtain an adaptive fill light network structure; Mapping the multimodal light source characteristics onto the adaptive fill light network structure to obtain a preliminary fill light strategy; The fill light source control pulse sequence is optimized based on the preliminary fill light strategy to obtain the fill light source adjustment strategy.

7. The method for adaptively controlling the supplementary light source of an industrial camera according to claim 6, characterized in that: Mapping the multimodal light source characteristics onto the adaptive fill light network structure to obtain a preliminary fill light strategy includes: Performing feature vector decomposition on the multimodal light source features to obtain principal components of light source features; Assigning node weights to the adaptive fill light network structure according to the principal components of the light source characteristics to obtain a weighted fill light network; Performing graph analysis on the weighted light-filling network and extracting network topology features to obtain a network structure feature vector; Performing a tensor product operation on the network structure feature vector and the light source feature principal component to obtain a fused feature tensor; Performing a nonlinear transformation on the fused feature tensor to obtain a candidate set of fill light strategies; Filtering the fill light strategy candidate set according to a preset fill light sorting standard to obtain a corresponding filtered fill light strategy set; The screened fill light strategy set is integrated to obtain the preliminary fill light strategy.

8. An adaptive control device for supplementary light source of an industrial camera, characterized in that: The method for adaptively controlling supplemental light sources for industrial cameras, as applied to any one of claims 1 to 7, comprises: An acquisition module is used to obtain scene perception data from an industrial camera, and to perform real-time classification of shooting scenes and environmental parameter analysis based on the scene perception data to obtain scene feature information; An analysis module is used to perform image quality evaluation on the scene feature information and compare it with a preset image quality standard to obtain an image quality matching degree; an association module, configured to perform control parameter analysis based on the image quality matching degree and the scene feature information to obtain initial fill light control parameters; a processing module, the processing module being used to acquire image data from the industrial camera, perform brightness distribution and detail feature analysis on the image data, and obtain corresponding image analysis results; A control module, configured to perform dynamic light source simulation on the scene feature information based on the image analysis result to obtain corresponding dynamic fill light source parameters; an execution module, configured to comprehensively process the initial fill light control parameters, the image analysis results, and the dynamic fill light source parameters to obtain a fill light source adjustment strategy; Based on the scene perception data, real-time classification of the shooting scene and environmental parameter analysis are performed to obtain scene feature information, including: Acquiring spectral data and depth data collected by the industrial camera and integrating them to obtain the scene perception data; Performing high-order singular value decomposition on the scene perception data and extracting cross-modal correlation features to obtain a modal coupling matrix; Performing scene construction based on the modal coupling matrix to obtain a scene semantic graph; Performing Betti number sequence calculation on the scene semantic graph according to a preset persistent homology algorithm to obtain a scene topology feature vector; Performing multi-scale fractal dimension analysis based on the scene topology feature vector to obtain a scene complexity spectrum; Performing dynamic stability feature analysis based on the scene complexity spectrum to obtain scene dynamic features; Performing symbolic processing on the dynamic features of the scene to generate a symbol sequence of the scene; Calculating conditional entropy and mutual information based on the symbol sequence to obtain a scene uncertainty index; The scene topology feature vector, the scene dynamic feature and the scene uncertainty index are subjected to nonlinear mapping fusion to obtain the scene feature information.

Images